Benjamini Hochberg
This program applies the Benjamini Hochberg correction to a column of P values. The significant P values are extracted and the adjusted Benjamini-Hochberg P values are computed as well. The FDR is 0.01 and the threshold is 0.05 by default.
Manual :
1- install Julia v1.6 or higher free programming language and install packages :
ArgParse : add ArgParse
CSV : add CSV
DataFrames : add DataFrames
StatsBase : add StatsBase
MultipleTesting : add MultipleTesting
2- Install Perl and Gnu parallel for parallel file processing
3- unzip the software
4- copy your n tables in csv (TAB delimitated by default) files in the “data” directory.
The tables must have at least one columns labeled Pvalue.
5- execute the software by the command :
julia BenjHochberg-011.jl [-s SEP] -f FILE [-d FDR] [-t THR] [-h]
optional arguments:
-s, --sep SEP separator for CSV tables (default: '\t')
-f, --file FILE file with at least one columns labeled Pvalue.
-d, --fdr FDR FDR ex. 0.01 (type: Float64, default: 0.01)
-t, --thr THR Qvalue Threshold ex. 0.05 (type: Float64, default: 0.05)
-h, --help show this help message and exit
example :
julia BenjHochberg-005.jl -f data/test.csv
julia BenjHochberg-0xx.jl -d 0.001 -t 0.05 -f data/test.csv # FDR = 0.001 significance threshold = 0.05
The FDR is 0.01 and the threshold is 0.05 by default.
To process files in parallel, edit the conf.txt file and start processing with the command perl parallel_BH-0.2.pl
6- processed file is in the “result” directory.
7- significant results are in "significant" directory.
for more explanations of Pvalue with FDR control, see here